Start with the right input
Text to video
Use text-to-video when the scene does not yet exist. Describe subject, action, setting, shot type, camera path and light in that order.
Image to video
Use image-to-video when composition, character design or product appearance is already established. The motion prompt should focus on change across time.
Multi-shot production
Plan a sequence as separate shots with one visual goal each. Reuse references and concise style language to maintain continuity.
AI video prompt formula
Keep negative constraints short. Conflicting camera or motion instructions can reduce coherence, so make one clear directing choice per shot.
Featured models
SuperMaker AI Video Agent
Chat- and brief-led end-to-end video workflow for coordinating image, motion, voice and music from one production plan.
Open model guide →02 / SuperMaker · Canvas StudioSuperMaker Canvas Workflow Studio
Node-based infinite canvas for Text → Image → Video pipelines, multi-model routing, batch variations and visual project history.
Open model guide →03 / Krea · Image DirectionKrea 2
An expressive foundation image model that helps production teams turn moodboards and style references into approved visual language.
Open model guide →